IAPP AIGP Practice Test 96
Question 950
An industrial plant needs near-real-time safety classification in an intermittently connected location with strict data controls. Postdeployment review has identified the same issue after a material service change. What action should the accountable team prioritize to govern this use responsibly?
- Pilot a small local model or simple algorithm when it meets safety and accuracy constraints.
- Prefer controlled retrieval of versioned policy sources for rapidly changing factual rules; fine-tune only for justified behavior goals.
- Deploy an appropriately validated edge or on-prem design with bounded synchronization when cloud reliance is unsafe.
- Define a site-readiness gate with infrastructure, workforce and performance acceptance criteria.
Explanation
Correct Answer
C:Deploy an appropriately validated edge or on-prem design with bounded synchronization when cloud reliance is unsafe. Correct. It directly reduces or governs the specific documented risk. AI deployment topology should match operational hazards and data obligations.
Incorrect Answers
A:Pilot a small local model or simple algorithm when it meets safety and accuracy constraints. Not selected. That action treats another control objective, not the specific gap described. It belongs with Small versus large model: Model size is one factor in suitability and can increase cost and exposure without adding needed value. The present case instead concerns Cloud edge on-prem choice.
B:Prefer controlled retrieval of versioned policy sources for rapidly changing factual rules; fine-tune only for justified behavior goals. Not selected. That action treats another control objective, not the specific gap described. It belongs with Fine-tuning versus prompt or RAG: Fine-tuning adapts model behavior but is not always suitable for rapidly changing facts. The present case instead concerns Cloud edge on-prem choice.
D:Define a site-readiness gate with infrastructure, workforce and performance acceptance criteria. Not selected. That action treats another control objective, not the specific gap described. It belongs with Deployment-context feasibility: The deployment environment and workforce can create new risk even when the model was sound in development. The present case instead concerns Cloud edge on-prem choice.
Question 951
An industrial plant needs near-real-time safety classification in an intermittently connected location with strict data controls. Postdeployment review has identified the same issue after a material service change. What result would provide the strongest case-specific assurance?
- Run field pilots at representative stores and review user comprehension and prediction failure rates.
- Test corrupted images, accented speech and conflicting text for safe handling and escalation.
- Measure tail latency, operational energy use, robustness and error rates on representative meters.
- Disconnect the network and verify local safe operation and recovery within defined thresholds.
Explanation
Correct Answer
D:Disconnect the network and verify local safe operation and recovery within defined thresholds. Correct. It yields measurable evidence that the case-specific safeguard works. AI deployment topology should match operational hazards and data obligations.
Incorrect Answers
A:Run field pilots at representative stores and review user comprehension and prediction failure rates. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Deployment-context feasibility: The deployment environment and workforce can create new risk even when the model was sound in development. The present case instead concerns Cloud edge on-prem choice.
B:Test corrupted images, accented speech and conflicting text for safe handling and escalation. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Multimodal model assessment: Multimodal AI adds failure channels that text-only benchmarks cannot substantiate. The present case instead concerns Cloud edge on-prem choice.
C:Measure tail latency, operational energy use, robustness and error rates on representative meters. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Small versus large model: Model size is one factor in suitability and can increase cost and exposure without adding needed value. The present case instead concerns Cloud edge on-prem choice.
Question 952
An industrial plant needs near-real-time safety classification in an intermittently connected location with strict data controls. Postdeployment review has identified the same issue after a material service change. Which TWO actions form the appropriate investigation-and-control pair for this case?
- Deploy an appropriately validated edge or on-prem design with bounded synchronization when cloud reliance is unsafe.
- Compare cloud, on-premises and edge deployment against latency, resilience, data location and incident-recovery needs.
- Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site.
- Pilot a small local model or simple algorithm when it meets safety and accuracy constraints.
Explanation
Correct Answers
A:Deploy an appropriately validated edge or on-prem design with bounded synchronization when cloud reliance is unsafe. Correct. As the second, implementation half of the requested pair, this directly applies the needed safeguard. AI deployment topology should match operational hazards and data obligations.
B:Compare cloud, on-premises and edge deployment against latency, resilience, data location and incident-recovery needs. Correct. As the first, evidence-gathering half of the requested pair, this establishes the specific risk boundary. AI deployment topology should match operational hazards and data obligations.
Incorrect Answers
C:Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site. Not selected. It addresses a different governance mechanism, not the required assessment-and-control pair. It belongs with Deployment-context feasibility: The deployment environment and workforce can create new risk even when the model was sound in development. The present case instead concerns Cloud edge on-prem choice.
D:Pilot a small local model or simple algorithm when it meets safety and accuracy constraints. Not selected. It addresses a different governance mechanism, not the required assessment-and-control pair. It belongs with Small versus large model: Model size is one factor in suitability and can increase cost and exposure without adding needed value. The present case instead concerns Cloud edge on-prem choice.
Question 953
An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. The operational owner must decide whether the current deployment can safely proceed. Before selecting a remedy, what should the accountable reviewer investigate first?
- Identify knowledge freshness, provenance, source authorization and citation requirements.
- Identify each input modality, potential cross-modal failure and missing evaluation coverage.
- Compare cloud, on-premises and edge deployment against latency, resilience, data location and incident-recovery needs.
- Run field pilots at representative stores and review user comprehension and prediction failure rates.
Explanation
Correct Answer
A:Identify knowledge freshness, provenance, source authorization and citation requirements. Correct. It establishes the relevant factual risk boundary before implementation. RAG can improve grounding but does not eliminate hallucination, leakage or access-control obligations.
Incorrect Answers
B:Identify each input modality, potential cross-modal failure and missing evaluation coverage. Not selected. That check primarily diagnoses a different governance issue. It belongs with Multimodal model assessment: Multimodal AI adds failure channels that text-only benchmarks cannot substantiate. The present case instead concerns Retrieval augmented generation.
C:Compare cloud, on-premises and edge deployment against latency, resilience, data location and incident-recovery needs. Not selected. That check primarily diagnoses a different governance issue. It belongs with Cloud edge on-prem choice: AI deployment topology should match operational hazards and data obligations. The present case instead concerns Retrieval augmented generation.
D:Run field pilots at representative stores and review user comprehension and prediction failure rates. Not selected. This action relates to Deployment-context feasibility. The deployment environment and workforce can create new risk even when the model was sound in development. For the diagnose decision, it does not resolve the specifically described issue of Retrieval augmented generation in An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. The operational owner must decide whether the current deployment can safely proceed.
Question 954
An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. The operational owner must decide whether the current deployment can safely proceed. What action should the accountable team prioritize to govern this use responsibly?
- Choose a suitable tested anomaly detector for bounded classification and reserve generative text for appropriate assistance.
- Require multimodal evaluation, input-integrity controls and restrictions on unsupported modality combinations.
- Use controlled retrieval from approved current documents with access filtering and source-aware answers.
- Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site.
Explanation
Correct Answer
C:Use controlled retrieval from approved current documents with access filtering and source-aware answers. Correct. It directly reduces or governs the specific documented risk. RAG can improve grounding but does not eliminate hallucination, leakage or access-control obligations.
Incorrect Answers
A:Choose a suitable tested anomaly detector for bounded classification and reserve generative text for appropriate assistance. Not selected. That action treats another control objective, not the specific gap described. It belongs with Classical versus generative choice: Deployment model type should follow the task and assurance requirements, not a trend toward generative AI. The present case instead concerns Retrieval augmented generation.
B:Require multimodal evaluation, input-integrity controls and restrictions on unsupported modality combinations. Not selected. That action treats another control objective, not the specific gap described. It belongs with Multimodal model assessment: Multimodal AI adds failure channels that text-only benchmarks cannot substantiate. The present case instead concerns Retrieval augmented generation.
D:Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site. Not selected. This action relates to Deployment-context feasibility. The deployment environment and workforce can create new risk even when the model was sound in development. For the control decision, it does not resolve the specifically described issue of Retrieval augmented generation in An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. The operational owner must decide whether the current deployment can safely proceed.
Question 955
An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. The operational owner must decide whether the current deployment can safely proceed. What result would provide the strongest case-specific assurance?
- Test version-changed policy questions and confirm citations, permissions and abstention behavior.
- Test corrupted images, accented speech and conflicting text for safe handling and escalation.
- Disconnect the network and verify local safe operation and recovery within defined thresholds.
- Run field pilots at representative stores and review user comprehension and prediction failure rates.
Explanation
Correct Answer
A:Test version-changed policy questions and confirm citations, permissions and abstention behavior. Correct. It yields measurable evidence that the case-specific safeguard works. RAG can improve grounding but does not eliminate hallucination, leakage or access-control obligations.
Incorrect Answers
B:Test corrupted images, accented speech and conflicting text for safe handling and escalation. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Multimodal model assessment: Multimodal AI adds failure channels that text-only benchmarks cannot substantiate. The present case instead concerns Retrieval augmented generation.
C:Disconnect the network and verify local safe operation and recovery within defined thresholds. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Cloud edge on-prem choice: AI deployment topology should match operational hazards and data obligations. The present case instead concerns Retrieval augmented generation.
D:Run field pilots at representative stores and review user comprehension and prediction failure rates. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Deployment-context feasibility: The deployment environment and workforce can create new risk even when the model was sound in development. The present case instead concerns Retrieval augmented generation.
Question 956
An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. The operational owner must decide whether the current deployment can safely proceed. What should the owner do to prevent further harm from this specific failure?
- Withdraw a generative solution that cannot reliably meet the deterministic decision criteria.
- Remove stale model-baked claims and fall back to verified source documents.
- Disable unsupported policy advice and provide verified staff assistance during remediation.
- Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site.
Explanation
Correct Answer
C:Disable unsupported policy advice and provide verified staff assistance during remediation. Correct. It limits further reliance or exposure while the risk remains unresolved. RAG can improve grounding but does not eliminate hallucination, leakage or access-control obligations.
Incorrect Answers
A:Withdraw a generative solution that cannot reliably meet the deterministic decision criteria. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with Classical versus generative choice: Deployment model type should follow the task and assurance requirements, not a trend toward generative AI. The present case instead concerns Retrieval augmented generation.
B:Remove stale model-baked claims and fall back to verified source documents. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with Fine-tuning versus prompt or RAG: Fine-tuning adapts model behavior but is not always suitable for rapidly changing facts. The present case instead concerns Retrieval augmented generation.
D:Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site. Not selected. This action relates to Deployment-context feasibility. The deployment environment and workforce can create new risk even when the model was sound in development. For the contain decision, it does not resolve the specifically described issue of Retrieval augmented generation in An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. The operational owner must decide whether the current deployment can safely proceed.
Question 957
An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. A new business unit wants to scale the system without a further governance review. What should the AI governance lead examine before changing the system?
- Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site.
- Compare fine-tuning, prompt design and retrieval on update frequency, provenance, cost and rights.
- Identify knowledge freshness, provenance, source authorization and citation requirements.
- Define a site-readiness gate with infrastructure, workforce and performance acceptance criteria.
Explanation
Correct Answer
C:Identify knowledge freshness, provenance, source authorization and citation requirements. Correct. It establishes the relevant factual risk boundary before implementation. RAG can improve grounding but does not eliminate hallucination, leakage or access-control obligations.
Incorrect Answers
A:Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site. Not selected. That check primarily diagnoses a different governance issue. It belongs with Deployment-context feasibility: The deployment environment and workforce can create new risk even when the model was sound in development. The present case instead concerns Retrieval augmented generation.
B:Compare fine-tuning, prompt design and retrieval on update frequency, provenance, cost and rights. Not selected. That check primarily diagnoses a different governance issue. It belongs with Fine-tuning versus prompt or RAG: Fine-tuning adapts model behavior but is not always suitable for rapidly changing facts. The present case instead concerns Retrieval augmented generation.
D:Define a site-readiness gate with infrastructure, workforce and performance acceptance criteria. Not selected. This action relates to Deployment-context feasibility. The deployment environment and workforce can create new risk even when the model was sound in development. For the diagnose decision, it does not resolve the specifically described issue of Retrieval augmented generation in An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. A new business unit wants to scale the system without a further governance review.
Question 958
An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. A new business unit wants to scale the system without a further governance review. What is the most directly applicable governance control for this situation?
- Use controlled retrieval from approved current documents with access filtering and source-aware answers.
- Pilot a small local model or simple algorithm when it meets safety and accuracy constraints.
- Require multimodal evaluation, input-integrity controls and restrictions on unsupported modality combinations.
- Deploy an appropriately validated edge or on-prem design with bounded synchronization when cloud reliance is unsafe.
Explanation
Correct Answer
A:Use controlled retrieval from approved current documents with access filtering and source-aware answers. Correct. It directly reduces or governs the specific documented risk. RAG can improve grounding but does not eliminate hallucination, leakage or access-control obligations.
Incorrect Answers
B:Pilot a small local model or simple algorithm when it meets safety and accuracy constraints. Not selected. That action treats another control objective, not the specific gap described. It belongs with Small versus large model: Model size is one factor in suitability and can increase cost and exposure without adding needed value. The present case instead concerns Retrieval augmented generation.
C:Require multimodal evaluation, input-integrity controls and restrictions on unsupported modality combinations. Not selected. That action treats another control objective, not the specific gap described. It belongs with Multimodal model assessment: Multimodal AI adds failure channels that text-only benchmarks cannot substantiate. The present case instead concerns Retrieval augmented generation.
D:Deploy an appropriately validated edge or on-prem design with bounded synchronization when cloud reliance is unsafe. Not selected. That action treats another control objective, not the specific gap described. It belongs with Cloud edge on-prem choice: AI deployment topology should match operational hazards and data obligations. The present case instead concerns Retrieval augmented generation.
Question 959
An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. A new business unit wants to scale the system without a further governance review. What should an independent reviewer test before signing off on this specific risk?
- Measure tail latency, operational energy use, robustness and error rates on representative meters.
- Validate data-flow boundaries, patchability, access logging and total operating cost for the chosen option.
- Test version-changed policy questions and confirm citations, permissions and abstention behavior.
- Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site.
Explanation
Correct Answer
C:Test version-changed policy questions and confirm citations, permissions and abstention behavior. Correct. It yields measurable evidence that the case-specific safeguard works. RAG can improve grounding but does not eliminate hallucination, leakage or access-control obligations.
Incorrect Answers
A:Measure tail latency, operational energy use, robustness and error rates on representative meters. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Small versus large model: Model size is one factor in suitability and can increase cost and exposure without adding needed value. The present case instead concerns Retrieval augmented generation.
B:Validate data-flow boundaries, patchability, access logging and total operating cost for the chosen option. Not selected. That test would not demonstrate this particular risk has been addressed. It belongs with Open versus proprietary model: Open or proprietary licensing changes governance tradeoffs but neither eliminates accountability. The present case instead concerns Retrieval augmented generation.
D:Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site. Not selected. This action relates to Deployment-context feasibility. The deployment environment and workforce can create new risk even when the model was sound in development. For the validate decision, it does not resolve the specifically described issue of Retrieval augmented generation in An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. A new business unit wants to scale the system without a further governance review.
Question 960
An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. A new business unit wants to scale the system without a further governance review. What is the most relevant fallback when the promised safeguard cannot be demonstrated?
- Switch to a controlled alternative when licensing or privacy assurances fail.
- Restore a lightweight validated method when the large model cannot meet edge constraints.
- Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site.
- Disable unsupported policy advice and provide verified staff assistance during remediation.
Explanation
Correct Answer
D:Disable unsupported policy advice and provide verified staff assistance during remediation. Correct. It limits further reliance or exposure while the risk remains unresolved. RAG can improve grounding but does not eliminate hallucination, leakage or access-control obligations.
Incorrect Answers
A:Switch to a controlled alternative when licensing or privacy assurances fail. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with Open versus proprietary model: Open or proprietary licensing changes governance tradeoffs but neither eliminates accountability. The present case instead concerns Retrieval augmented generation.
B:Restore a lightweight validated method when the large model cannot meet edge constraints. Not selected. That response contains a different failure mode and leaves the stated risk unresolved. It belongs with Small versus large model: Model size is one factor in suitability and can increase cost and exposure without adding needed value. The present case instead concerns Retrieval augmented generation.
C:Assess business objectives, data availability, operator skills, network conditions and failure impact at each deployment site. Not selected. This action relates to Deployment-context feasibility. The deployment environment and workforce can create new risk even when the model was sound in development. For the contain decision, it does not resolve the specifically described issue of Retrieval augmented generation in An insurer's general-purpose assistant invents current policy terms because its base model has no access to the latest approved documents. A new business unit wants to scale the system without a further governance review.